First Progress Report - 6. to plan and commence tradescantia control for best results in the given sites
Initial findings in mid-March in the Tradescantia zones chosen for this Trial were:
1) Except at the very outer margin adjacent mown exotic grasses and herbs, no seedlings, either weed or native, were present in the tall deep Tradescantia among the planted trees, ie from the road edge to the 1997 margin of wild regeneration that bordered much, but not all, of the stream on the Kaipatiki Rd side.
2) Dessication of the soil during late summer drought was severe, except in the area of our earlier, October 2018, exploratory weeding of tradescantia, which was damp beneath the tradescantia. An area several metres wide appears to be a run off channel, its source unknown as the road above appears to be guttered here as it is throughout the length of the road.
Though since ameliorated in places by recent rains, the dry conditions persist (now early May) along the road edge, except in similar small areas of persistent run-off from unknown sources.
The removal of gutters the length of the road would allow rain to reach the soil alongside, but presumably no practical way has yet been found to remove gutters or make them permeable without massive erosion by the force of water collected during heavy rain events from the impermeable expanse of the wide road.
Most of the trees in the Tradescantia Trial site were planted in 1999 or shortly after. Those that have survived are presumably adapted to the moisture levels that prevail here, but some at least do appear to be sensitive to sudden loss of ground cover during drought, and, should ground cover loss occur as it did in our initial removal of tradescantia and of dumped refuse, to benefit from its replacement with loose tradescantia.
We began exploratory removal of Tradescantia in the Trad zones Ca, Cb, CC, Da and Db on 17th March, after a day of steady rain. All appeared fresh and lush, including the apparently dense Tradescantia 40-50cm high.
Surprisingly, when uplifted the sward was found to be composed of fewer than expected, widely-separated stems, 1-2m long, shallowly rooted and easily removed entire. The soil surface had a coarse, permeable, fibrous texture, with scant leaf litter and humus. The superficial root hairs of the many mahoe were presumably obtaining moisture from the very surface of the soil, shaded and kept moist by the Tradescantia which by its horizontally-layered arrangement presented a continuous barrier to evaporation.
Also observed on disturbing the Tradescantia were S. African mantis (2 ) and a Giant centipede. It scuttled to safety too fast to photograph, but the location was interesting:
https://inaturalist.nz/observations/21591744
We observed leaf droop among mahoe and kawakawa when noonday sun shone through the light and incomplete mahoe canopy on the day after removing 100% ground cover by inorganic refuse, and about 20% of the ground cover by Tradescantia, from this upper bank during drought.
This leaf droop extended over the next week to mahoe and some kawakawa in the less-dry area where we had removed ground cover by Tradescantia from approx 25% of the area, with no significant refuse removal.
We were able to restore ground cover throughout within a day or so of the initial observation of leaf droop, by gathering additional Tradescantia from the lower bank, but in several trees in the area of the refuse removal the leaf droop continued, then worsened. Some trees were affected and some not. Juveniles and adults appeared similarly affected. On one large mahoe, outer foliage at the roadside was unaffected while leaves under the canopy drooped. This we attributed possibly to the exposed foliage being supplied by roots in the deeper humus created by the roaqdside weeds ongoing development and decomposition, or possibly to Tradescantia having been retained at the road edge during refuse removal in order to suppress the exotic grasses there.
Further exploration revealed that this Western end of the Trial site, as well as the bank beyond it which had received no intervention, was drier throughout, indicated among other things by purpling of the tradescantia beyond the Trial site, where Tradescantia was also less dense and tall.
About 10 days after dessication by refuse and Tradescantia removal, during which a light rain produced no obvious improvement in the affected trees, there was another rain, heavy at times, and the turgidity of the mahoe leaves was partially restored.
A subsequent day of light rain within the next week returned appearances to normal throughout the site.
At the base of some trees near the top of the upper bank the soil remained completely dry, even during light rainfall; presumably these tree bases are canopied by their own or other foliage, preventing light-moderate rain reaching the ground.
So Tradescantia removal needs to be staged very cautiously, and be accompanied by replacement ground cover by mulch eg decomposed tradescantia. The entire infestation could have been removed entirely while the ground was dry, and would not have returned withn that area, only by reinvasion from adjacent infestation. Seedlings and juveniles being absent on those upper banks, it would have been a quick job , and an opportunity to demonstrate the effectiveness of manual removal. However, we believe it would have been followed not by seedling development, but by the permanent exposure of existing tree roots surrounded by dry, bare compacted clay, and the permanent loss of many other biota currently supported by the Tradescantia.
This effect has been seen in a home "garden" (wild regen native forest margin from which a moderate invasion of Tradescantia was removed between 1999-2001, and remains tradescantia-free (though all neighbouring properties support it to the fenceline, where it is controlled by hand), and has produced few seedlings of any kind in the subsequent years .
The effect was also seen in the downstream Kaipatiki Rd bank, where Tradescantia 100% removed by hand in 1999, and the development of native seedlings in the margin of wild regen canopy seems to have stopped at just that point. There is even a staked 1m quadrant in which a student conducted a seedling count. (We were not informed if the study was concluded). That spot is now a dry, mostly bare clay bank ,with a few strands of Tradescantia making a slow approach from the margin of the nearest current infestation. However, we have not been able to ascertain whether or when any firther weed control interventions occurred. It would be very useful to the assessment of manual removal to have information on any other interventions in the area, whether manual or chemical
.
https://inaturalist.nz/observations/17871842
and https://inaturalist.nz/observations/16582848
and an area nearby unweeded due to steepness of bank:
https://inaturalist.nz/observations/16619306
Winter rains should ameliorate the ground conditions in due course, but we need to anticipate similar conditions next summer. Therefore a very limited removal of Tradescantia is planned for this Trial. It will be along the wetter lower bank, "KRS", from where it drops steeply to the stream. It will also be removed from this steep bank directly above the stream wherever moisture is adequate, native regeneration likely, and access possible.
KRS bank will provide some useful data as we monitor seedling and sporeling generation in this limited area, including weed seedlings which are currently absent frm the Tradescantia-covered banks despite many mature fruiting tree privet, woolly nightshade, brush wattle, Elaeagnus, Kahili ginger, Japanese honeysuckle, mothplant and Calystegia all around them and overhead. In the proximity of nikau and other kereru-feeding trees, we also expect Tradescantia release to produce Phoenix and Bangalow palm seedlings, as they are now common in streamside areas without dense ground cover
We also hope to demonstrate complete removal from some small areas of differing conditions
On the porous sand banks of the stream itself, densely covered in lush understorey, a small area has been weeded already. A few native seedlings and sporeling have been released during this initial exploration, and further development will be monitored
Due to the nature of stream bank and dam formation, much of the Tradescantia is rooted deeply in many layers of sand, so there will be ongoing regrowth, along with reinvasion through flooding from upstream, with entrapment among the logs and debris that form dams.
Comments
Add a Comment